The Role Of The Endocannabinoid System In Gametogenesis, Implantation

Thread starter #1

Maternal use of marijuana, in which the exocannabinoid Δ9-tetrahydrocannabinol is the most active psychoactive ingredient, is known to have adverse effects on various aspects of reproduction including ovulation, spermatogenesis, implantation and pregnancy duration. Endogenous cannabinoids of which Anandamide is the prototype are widely distributed in the body especially in the reproductive tract and pregnancy tissues and act through the same receptors as the receptor as Δ9-tetrahydrocannabinol. Anandamide, has been reported to have pleiotropic effects on human reproduction and in experimental animal models. It appears to be the important neuro-cytokine mediator synchronizing the embryo-endometrial development for timed implantation, the development of the embryo into the blastocyst and transport of the embryo across the fallopian tubes. The mechanisms by which it exerts these effects are unclear but could be via direct actions on the various sites within the reproductive system or its differential actions on vascular tone dependent. In this review article we bring together the current knowledge on the role of endoccanabinoids in reproduction and postulate on the potential mechanisms on how these affect reproduction. In addition, we examine its role on the endothelium and vascular smooth muscle as a potential mechanism for adverse pregnancy outcome.


Pregnancy complications such as preterm labour, pre-eclampsia and fetal growth restriction (FGR) collectively make a significant contribution to perinatal morbidity and mortality. Although the pathophysiology has not been clearly defined, in most cases, the common phenomenon observed between these diseases is abnormal development and function of the placenta (Salafia, 1997; Pardi et al., 1997; Kim et al., 2002; McMaster et al., 2004).

Normal placental development is dependent upon the differentiation and invasion of the trophoblast, the main cellular component of the placenta that originates from the trophoectoderm of the blastocyst in early pregnancy (Straszewski-Chavez et al., 2005). During this process of development and invasion, trophoblast cells rapidly divide to form the interface between mother and embryo. Other trophoblast subpopulations invade the decidua to remodel the uterine spiral arteries allowing the expansion of extra-embryonic tissues and increase in blood flow to the placenta and developing fetus. Any perturbation of this process, which is tightly regulated and influenced by several factors, may lead to pregnancy complications. Such factors include numerous angiogenic growth factors, cell adhesion molecules, cytokines and growth factors, extracellular matrix metalloproteases, hormones and transcription factors which have been studied extensively (Vuorela et al., 1997; Kayisli et al., 2002; Rajashekhar et al., 2005; Walter and Schonkypl, 2006). Another family of bioactive factors that have not been studied in depth, but thought to be involved in normal placentation is the endocannabinoids.

In the light of recent evidence and increased interest in the role of endocannabinoids in early and late pregnancy problems, this review will re-examine the role of cannabinoids in pregnancy and in the development of the fetoplacental unit with special focus upon the effects of endocannabinoids on the endothelial and vascular smooth muscle cell as pertaining to efficient placental function.

The endocannabinoid system

Historical view
Research on the chemistry and pharmacology of cannabinoids was well underway in the 19th century, during a time when cannabis was widely used in medicine (Mechoulam and Hanus, 2000). However, it was more than a century later, that its most psychoactive active component, Δ9-tetrahydrocannabinol (Δ9-THC), was isolated in its pure form and its chemical structure elucidated (Gaoni and Mechoulam, 1964, 1971). Δ9-THC (Fig. 1) is the principal biologically active component of marijuana and is the prototypical cannabinoid family member. This and related molecules (Δ8-THC, Fig. 1) were originally shown to exert their central psychoactive effects via the cannabinoid receptor, CB1, and their peripheral immunoregulatory effects via the CB2 receptor (Matsuda et al., 1990; Munro et al., 1993). Recently, it has been shown that some peripheral tissues contain both receptor isotypes and, in some cases, Δ9-THC may also activate the nociceptive receptor, vanilloid receptor 1 (VR1) (Zygmunt et al., 2000; Zygmunt et al., 2002).

Endocannabinoids are generated 'on demand' from long chain polyunsaturated fatty acid precursors derived from arachidonic acid (Habayeb et al., 2002), through enhanced intracellular Ca2+ concentrations, e.g. from cell depolarization, or mobilization of intracellular Ca2+ stores following stimulation of Gq/11 protein-coupled receptors. The enzymes (N-arachidonyol-phosphatidyl ethanolamine selective phospholipase D and sn-1 selective diacylglycerol lipases 1 and 2) that catalyse the last steps in the production of the two most studied endocannabinoids, arachidonyol-ethanolamine (anandamide, AEA, Fig. 1) and 2-arachidonyl-glycerol (2-AG, Fig. 1) are all Ca2+-sensitive. It is widely accepted that AEA is generated in most cell types by the hydrolysis of the membrane phospholipid precursor, N-acylphosphatidylethanolamine (NAPE) by an enzyme that belongs to the phospholipase D family (Di Marzo et al., 1996a; Wang et al., 2006a,b), although very recent evidence suggests that in some cases, AEA can be generated through a phospholipase C-dependent pathway (Liu et al., 2006) or through an α/β-hydrolase 4 enzymatic system (Fig. 2; Simon and Cravatt, 2006).

After initial generation within the cell, the endocannabinoids are released into the interstitial space where they either diffuse into the lymph and blood or act in an autocrine or paracrine manner (Piomelli et al., 2000). The mechanism whereby endocannabinoids are released and which cells actually produce the compound is a matter of intense research (Di et al., 2005; Jung et al., 2005; Vogeser et al., 2006). Platelets and endothelial cells are one source of these compounds (Maccarrone et al., 2000a), although the ubiquitous distribution of arachidonic acid and the enzymes that generate NAPE suggest that other cell types including neurons may be the primary source (Liu et al., 2006). However, in the context of reproduction, evidence that estradiol (E2) enhances and progesterone inhibits AEA release from endothelial cells (Maccarrone et al., 2002a) suggests that the endothelial cell could be a prime target for the control of endocannabinoid production in the developing placenta, since the trophoblast progressively produces progesterone which would inhibit AEA release by endothelial cells. By contrast, the denudation of the endothelial layer in maternal spiral arteries that occurs during the second phase of placental development (Redman and Sargent, 2005) is regulated by the invading cytotrophoblast (Cartwright et al., 2002), a process that is under both estrogen and progesterone control. The sudden increase in local AEA concentrations at the invading tips could be a signal for local endothelial cell apoptosis in the decidua and the signal for smooth muscle apoptosis in the myometrium, although more work in this area is required (Thadani et al., 2004). Furthermore, the presence of AEA in tissue macrophages (Di Marzo et al., 1996b) and within many reproductive fluids including semen (Schuel et al., 2002a) suggests that other gonadal steroid target tissues and cells may also be endocannabinoid producers.

Metabolism–transport and degradation
Termination of endocannabinoid signalling at the cannabinoid receptors is thought to occur by transport of the compounds into the cell by a poorly characterized AEA transporter (Di Marzo et al., 2004; Moore et al., 2005; Bari et al., 2006), followed by the rapid degeneration of AEA by fatty acid amide hydrolase (FAAH) found on the internal membranes of AEA target cells (Di Marzo et al., 1994; Cravatt et al., 1996) (Figs. 2 and 3). It has also been suggested that FAAH may act as an AEA transporter and that FAAH not only absorbs AEA from the plasma, but may also export AEA under the appropriate conditions, i.e. estrogen stimulation of loaded endothelial cells (Maccarrone et al., 2002a). The nature of the AEA/FAAH transport molecule in the human has recently been suggested to be an alternative isoform of the original FAAH enzyme (Wei et al., 2006) found on the internal cell surface (Cravatt et al., 1996; Giang and Cravatt., 1997). This novel isoform (type 2) is found on the external surface of the cell (Wei et al., 2006). However, mice appear to express only one isoform of FAAH suggesting that the human and mouse utilize different mechanisms to modulate AEA levels (Wei et al., 2006). There is, however, compelling evidence that AEA is eliminated by transport into and subsequent degradation by intracellular FAAH from human umbilical vein endothelial cell (HUVEC) (Maccarrone et al., 2000a) and human peripheral T lymphocyte studies (Maccarrone et al., 2003a; Maccarrone et al., 2003b). From the HUVEC studies, it is clear that the AEA and FAAH transporters are indeed two independent entities, however, their precise chemical nature is yet to be defined. The currently held view is that the levels of FAAH in T lymphocytes are the sole method of regulating plasma AEA levels despite there being considerable evidence for FAAH expression in sites other than peripheral T lymphocytes (Maccarrone et al., 2003a; Park et al., 2003). The localization of FAAH in human amniotic epithelial cells, chorionic cytotrophoblastic cells, placenta and mouse endometrial epithelial cells (Maccarrone et al., 2002c; Park et al., 2003) and the human uterus (Helliwell et al., 2004) indicates a potentially important role in controlling AEA tone or attenuating local AEA-mediated events. The activity of uterine FAAH is regulated by sex hormones (Maccarrone et al., 2001, 2002c), suggesting that the endocannabinoid system may be involved in the maintenance of human pregnancy. Furthermore, in both in vivo and in vitro studies, the presence of FAAH in maternal lymphocytes has been shown to be integral in pregnancy continuation, as its down-regulation is an early marker for spontaneous abortion (Maccarrone and Finazzi-Agro, 2004). It has also been demonstrated that T-lymphocytes increase their levels of active FAAH approximately two-fold in response to exogenous progesterone (Maccarrone et al., 2001), supporting a primary role for FAAH in regulating AEA plasma levels during the first trimester of pregnancy and possibly local AEA level regulation.

Endocannabinoid receptors
Cannabinoid receptors

Two cannabinoid receptors (CB1 and CB2) have been identified and cloned from several species (Pertwee, 1997; Chataigneau et al., 1998), and a third (CB3) has been proposed, but has yet to be cloned (Fride et al., 2003). In most species, the major receptor, CB1, is primarily expressed in the cerebrum (Matsuda et al., 1990) and cerebellum (Kishimoto and Kano, 2006) but has also been demonstrated in peripheral tissues such as the eye, placenta, the fetal membranes and myometrium (Straiker et al., 1999; Park et al., 2003; Dennedy et al., 2004). The CB2 receptor, on the other hand, is mainly expressed by cells of the immune system (Munro et al., 1993), although transcripts for this gene have been identified in the placenta and trophoblastic cells (Buckley, et al., 1998). Distinction between these receptors is based on differences in their amino acid sequence (Matsuda et al., 1990; Munro et al., 1993), signalling mechanisms (Felder et al., 1995), tissue distribution (Matsuda et al., 1990; Munro et al., 1993; Shire et al., 1995), sensitivity to certain agonists such as WIN 55212-2 (a pravadoline derivative of the aminoalkylindole group) (Martin et al., 1991; Pertwee 1997), 1-propyl-2-methyl-3(1-naphthoyl)indole and 2-methylarachidonyl-(2′-fluoroethyl)amide (Huffman et al., 1996; Showalter et al., 1996) and antagonists that show receptor selectivity such as SR141716A and LY320135 (Compton et al., 1996; Felder et al., 1998). CB1 receptors have been shown to be involved in the relaxant and vasodilator actions of AEA via the antagonistic action of the CB1 receptor antagonist, SR141716A (Randall et al., 1996; Zygmunt et al., 1997). However, adding to the complexity of cannabinoid action, other studies have demonstrated that the vasodilatory response to AEA (in rat mesenteric arteries) may involve two other receptors: a non-CB1 endothelial receptor that is also sensitive to the effects of SR141716A and another vascular smooth muscle receptor i.e. resistant to SR141716A (Wagner et al., 1999). The relative contribution of these two components to the net response to AEA is thought to depend on the species and tissue used and on the condition of the endothelium, which may explain why conflicting studies have demonstrated both the ability and the inability of SR141716A to inhibit the vasorelaxant effect of AEA (Randall et al., 1996; Wagner et al., 1999). Furthermore, in rat mesenteric arteries, SR141716A was shown to inhibit vascular responses to the endothelium-dependent vasorelaxants, carbachol and the calcium ionophore, A23187 (Randall et al., 1996; White and Hiley 1997), whereas other studies have failed to demonstrate similar responses in guinea-pig carotid, rat mesenteric, porcine coronary and bovine coronary arteries (Plane et al., 1997; Pratt et al., 1998).

Finally, although many studies have demonstrated the action of AEA via cannabinoid receptors, a putative endothelial non-CB1 non-CB2 receptor has also been suggested by studies demonstrating a persistence in AEA-induced vasodilatation in mice deficient in both CB1 and CB2 receptors (Jarai et al., 1999). This has led to the idea that a third CB receptor may exist, and, indeed, the in silico analysis of CB gene structure indicates that GPR55 may be considered a strong candidate for the CB3 receptor (Baker et al., 2006).

Vanilloid receptors

Structural similarities between AEA and ligands of the capsaicin-sensitive vanilloid receptors (VR1) that signal and control nociception indicate that AEA may produce some of its effects through a low-affinity interaction with vanilloid receptors (VR1) (Di Marzo et al., 1998; Zygmunt et al., 1999). However, such (in vitro) effects have been shown to require supra-physiological concentrations of AEA, suggesting that AEA may act in concert with VR1-stimulating activity, leading to the idea of endovanilloids (Ross, 2003; Van Der Stelt and Di Marzo, 2004). These findings reinforce the complexity of AEA activity, because in addition to the involvement of CB1 receptors in some blood vessels, it has an SR141716A-sensitive component mediated by an as yet unidentified endothelial receptor and an endothelium-independent, SR141716A-resistant component likely mediated by vanilloid receptors (Jarai et al., 1999).

Role of cannabinoids in reproduction

To the best of our knowledge, there have been very few published studies on the effects of AEA on folliculogenesis. Most of the published studies have examined the effects of the active components of marijuana, namely, Δ9-THC, on oocyte development. [It is worth noting that older papers refer to Δ1-THC, which is in fact the same molecule as Δ9-THC.]

Δ9-THC has been shown to inhibit ovulation by suppressing plasma follicle-stimulating hormone (FSH) and the pre-ovulatory surge of LH when administered to rats on the day of proestrus (Nir et al., 1973; Ayalon et al., 1977). Reich et al. (1982) suggested that this may be primarily due to the hypothalamic inhibition of GnRH secretion. Furthermore, the inhibition of ovarian prostaglandin synthesis by Δ1-THC has also been demonstrated in granulosa cells, suggesting a direct suppressive effect of the drug on the ovary (Lewysohn et al., 1984). Furthermore, Δ9-THC has also been shown to cause a dose-dependent inhibition of the FSH-stimulated accumulation of progesterone and estrogen in ovarian granulosa cells (Lewysohn et al., 1984).

Although there is a paucity of data concerning the actions of endocannabinoids on the oocyte, AEA and its congeners, N-palmitoylethanolamine and N-oleoylethanolamine have been quantified in follicular fluid retained following oocyte aspiration from women undergoing IVF treatment (Schuel et al., 2002a). These endocannabinoids are thought to be produced by the granulosa cells in ovarian follicles and in granulosa cells surrounding ovulated oocytes. However, the mechanism controlling their production and release are currently unknown but are presumably under hormonal control. Additionally, cannabinoid receptors have been demonstrated to be present in the ovary (Galiegue et al., 1995), suggesting a local paracrine action of endocannabinoids. Ovulation is dependent upon cAMP accumulation, an effect that is inhibited by Δ9-THC in cultured rat granulosa cells (Treinen et al., 1993), suggesting that endocannabinoid signalling may help regulate follicle maturation and development (Schuel et al., 2002b). It may be this mechanism that accounts for the reported adverse effects of marijuana on ovulation (Abel, 1981; Powell and Fuller, 1983; Mueller et al., 1990; Schuel et al., 2002a). These data suggest that endocannabinoids exert both a direct and an indirect effect on ovulation.

Another possible additional mechanism for an indirect effect of endocannabinoids on folliculogenesis is via modulation of nutritional status. There is some evidence linking obesity, leptin production and reproduction (Clarke and Henry, 1999; Messinis and Milingos, 1999; Goumenou et al., 2003; Bajari et al., 2004; Linne, 2004) with obesity having an adverse effect on reproductive potential, and weight loss prior to assisted reproduction treatment having a beneficial effect on pregnancy outcomes (Fedorcsak et al., 2004). Recently, Di Marzo et al. (2001) elegantly linked the process of food intake, leptin production and endocannabinoids in the homozygous and heterozygous CB1-knockout mouse models, whereby defective leptin signalling leads to an increased hypothalamic cannabinoid level and action. In this way, it was proposed that an inhibitory action on pituitary gonadotrophin release is produced which has an adverse effect on ovulation. However, although the inverse relationship between AEA and leptin is clearly demonstrated in those studies as far as appetite and feeding is concerned, a direct link between these two molecules in the control of reproduction has not yet been found (Mechoulam and Fride, 2001).

Additional evidence in the control of appetite and obesity has come from clinical trials with the CB1 antagonist, SR141716 (Rimonabant), with several reports of successful regulation of body weight (Van Gaal et al., 2005; Cleland et al., 2004) and significant reductions in body mass in type 2 diabetics (Hollander, 2007). Although these results are encouraging as regards appetite suppression and body mass control, the use of CB1 antagonists in this way must be treated with caution since the precipitation of multiple sclerosis in pre-disposed individuals from the use of this drug may occur (van Oosten et al., 2004). Additionally, there are no data from these trials about reproductive function and whether the use of Rimonabant prior to assisted reproduction procedures would have any beneficial or adverse effects. Indeed, obesity control through the regulation of the endocannabinoid system requires much more research.

Although it is well known that chronic marijuana use transiently decreases male fertility in animal models and humans (Murphy et al., 1994), there are relatively few published studies on Δ9-THC and human spermatogenesis. The mechanism involved in marijuana-induced infertility remains unclear, although several studies have implicated reduced testosterone secretion (Wenger et al., 2001; Kolodny et al., 1974), sperm production (Nahas et al., 2002; Dalterio et al., 1977; Patra and Wadsworth, 1990), sperm motility (Zimmerman et al., 1979; Ambrosini et al., 2003; Schuel and Burkman, 2005) and sperm viability (Schuel and Burkman, 2002b) as possible culprits. Additionally, chronic marijuana use is associated with reduced LH production both centrally and within the testes, suggestive of both central and local endocannabinoid effects, as is the case in the female. The local effect of cannabinoids appears to be mediated through a CB2-dependent mechanism that involves modulation of Sertoli cell FAAH expression and regulation of AEA levels (Maccarrone et al., 2003c), with the suggestion that the pro-apoptotic nature of AEA (Maccarrone et al., 2000c; Maccarone and Finazzi-Agro, 2003) is responsible for aged Sertoli cell eradication. However, this is only part of the story, since AEA action on Sertoli cells is age-dependent and in nascent cells is probably not mediated via the CB2 receptor, because binding of AEA to the CB2 receptors on these cells has an anti-apoptotic effect (Maccarrone et al., 2003c). In spite of this, there is increasing Sertoli cell survival in spermatogenesis which may be due to the FSH-dependent increase in Sertoli cell FAAH activity that further degrades local AEA levels, thereby preventing Sertoli cell apoptosis (Maccarrone et al., 2003c). This suggests that endocrine regulation of FAAH activity in the Sertoli cell and the subsequent hydrolysis of AEA is a major check-point in human male fertility and that relative activities of the CB receptors during early spermatogenesis are controlling factors.

The normal mature human sperm has been shown to respond to AEA via the CB receptors with a resultant reduced motility (Suarez and Ho, 2003; Rossato et al., 2005), reduced capacitation ability (Rossato et al., 2005) and reduced binding to the zona pellucida. In addition, there is a reduction in the hydrolysis of the oocyte membranes (Rossato et al., 2005), suggesting that excess AEA production in the female reproductive tract reduces the success of fertilization. Using the boar sperm as a model system, Maccarrone et al. (2005) have demonstrated that the acrosomal reaction of sperm requires a small amount of AEA that acts through first the CB1 and then the VR1 receptor for full activity, and it has been postulated that during sperm capacitation, a dual stage-dependent endocannabinoid effect is in operation where AEA in seminal plasma and uterine fluids prevents capacitation in freshly ejaculated sperm, whereas sperm that have travelled through the uterus and into the Fallopian tube are exposed to gradually reduced AEA concentrations that release the molecular 'brake' of CB1 inhibition so that capacitation occurs (Schuel et al., 2002b). These data suggest that a full understanding and subsequent treatment of male infertility requires better appreciation of the involvement of the endocannabinoid system in both the male and female reproductive tracts.

Fertilization and oviductal transport
During early pregnancy, the development of the pre-implantation embryo and their timely oviductal transport into the uterus occurs simultaneously. Both endogenous and exogenous cannabinoids have been demonstrated to inhibit embryo development at the 2-cell stage through a CB1-dependent mechanism (Paria et al., 1995, 1998). Although early embryos are reported to express CB2 transcripts (Sharov et al., 2003), the role and function of CB2 is currently unknown with the implication that CB2 is confined to the inner cell mass that develops into the fetus and the CB1 receptor expressed in the developing trophoblast (Wang et al., 2006a,b). Although the authors suggest that the CB1 receptor is the functional receptor for normal embryo growth and development and that the CB2 receptor is responsible for controlling stem cell populations (Sharov et al., 2003; Wang et al., 2006a,b), these data need further clarification.

Embryos reaching the early blastocyst stage will have developed and differentiated to the point of having gained implantation potential. It is now clear that the endocannabinoid system is also involved in embryo transport, with clear evidence from CB1 and CB2 knockout studies that oviductal transport is a CB1-dependent mechanism in mice (Paria et al., 2001; Wang et al., 2004). CB1 knockout mothers retain their embryos in the oviduct with about 40% of CB1 knockout mothers failing to implant their embryos through this mechanism (Paria et al., 2001). This suggests that CB1 receptor deficiencies in some women may have a role to play in tubal pregnancy or female infertility. Interestingly, mice treated with either the non-hydrolysable AEA analogue, methanandamide or Δ9-THC showed pregnancy loss with embryos retained in the oviduct (Wang et al., 2004), suggesting that either inhibited or enhanced cannabinoid signalling impairs embryo transport. From these data it is clear that appropriate CB1 expression and function in the Fallopian tube is a limiting factor in oviductal transport and pregnancy success.

The highest concentration of AEA found in any species thus far studied was in the non-implantation site of the pregnant mouse uterus, suggesting an important role in mammalian reproduction (Schmid et al., 1997). Indeed, these data have been confirmed and extended to indicate that the spatial and temporal expression of the N-arachidonylphosphatidylethanolamine-hydrolysing phospholipase D (NAPE-PLD) enzyme in the mouse uterus is the main determinant of local AEA concentrations (Guo et al., 2005) and that AEA in concentrations above a critical level is toxic to the implanting blastocyst (Wang et al., 1999).

Wang et al. (1999) also suggested that site-specific levels of AEA in the uterus may regulate implantation by promoting trophoblast differentiation at the sites of blastocyst implantation, while at the same time limiting invasion beyond those sites. Furthermore, levels of AEA in the implantation sites have been found to be lower compared with those at the inter-implantation sites, suggesting that the implanting blastocyst may influence the local levels of AEA within the mouse uterus (Schmid et al., 1997; Paria et al., 1999).

Through studies like these it is now clear that AEA may be considered to be involved in the regulation of the 'window' of implantation whereby embryonic development is synchronized with the preparation of the uterus for implantation (Schmid et al., 1997). The finding that higher levels of AEA in non-implantation sites and lower levels within implantation sites (Paria and Dey, 2000) suggests that high levels of AEA may be responsible for inhibition of trophoblastic proliferation, whereas low levels are required for trophoblastic proliferation (Paria and Dey, 2000). Furthermore, endocannabinoids also appear to play a major role in regulating the development of the blastocyst to guarantee successful implantation in the endometrium (Wang et al., 2003).

Fetal development
Several groups have investigated the role of AEA and FAAH as important signals in early fetal development in both human and animal models (Wang et al., 1999, 2000; Maccarrone and Finazzi-Agro, 2004). It has been reported that AEA is synthesized within the reproductive tract of the female mouse, with uterine AEA and blastocyst CB1 receptor levels significantly higher than those in the (mouse) brain, highlighting the importance of AEA in the early stages of pregnancy (Yang et al., 1996). AEA has also been shown to act on cannabinoid receptors expressed on the embryonic cell surface to regulate the development of the preimplantation embryo in mice (Paria et al. 1996).

Other studies in mice have demonstrated that exposure of early embryos to high levels of cannabinoids inhibits blastocyst formation, zonal hatching and trophoblastic growth and that these effects are mediated via CB1 receptors (Paria et al., 1995; Yang et al., 1996; Schmid et al., 1997; Wang et al., 1999, 2003). The developmental arrest that occurs in mouse blastocysts in a non-receptive uterine environment associates well with its higher levels of AEA (Paria et al., 1996; Schmid et al., 1997). However, these effects appear to vary depending on the concentration of AEA and the developmental stage of the embryo. At lower doses of AEA, blastocyst formation is inhibited while trophoblastic outgrowth is accelerated in vitro. In stark contrast, when exposed to higher doses of AEA, trophoblastic outgrowth within the blastocyst was significantly inhibited (Wang et al., 1999). As uterine AEA levels vary depending on the stages of uterine receptivity and non-receptivity, these observations suggest that the effects of AEA are biphasic, and differentially executed depending on the embryonic stage and the cannabinoid levels to which the fetus is exposed (Wang et al., 1999; Paria et al., 2002).

In several animal studies, Δ9-THC has been shown to adversely affect the course and outcome of pregnancy by retarding embryo development, resulting in fetal abnormalities and teratological malformations in the newborn (Paria and Dey, 2000). Although gross teratological malformations are unknown in women subjected to chronic marijuana use, an association between chronic marijuana smoking and spontaneous abortion has been demonstrated (Lockwood, 2000; Nygren and Andersen, 2001). Furthermore, a decrease in both the birthweight and birthlength of the newborn associated with a symmetrical pattern of FGR has also been reported in marijuana users (Frank et al., 1990). The symmetrical nature of the FGR suggests an early onset, most likely from the time of implantation or during placental development. It is thought that 'isolated' FGR may be a consequence of abnormal placental vascular adaptation to pregnancy, resulting in a high pressure, low volume flow system (Takagi et al., 2004; Thaete et al., 2004). The precise mechanisms for maintaining low vascular reactivity within both the maternal and placental vascular beds during healthy pregnancies are unknown, although, sex steroids and locally produced vasoactive factors are thought to be implicated (Jaffe 1983; Gangula et al., 1997; Lockitch, 1997; Sladek et al., 1997). Vasoactive substances can induce either vasodilation or vasoconstriction by acting directly on vascular smooth muscle or indirectly via the vascular endothelium (Ang et al., 2002). Additionally, our demonstration that plasma AEA levels decline during normal pregnancy indicates that endocannabinoids may also be involved in the regulation of on-going pregnancy (Habayeb et al., 2004), but there is no epidemiological evidence in the literature that AEA or Δ9-THC adversely affect in utero fetal development, although there is evidence that marijuana use during pregnancy affects postnatal infant behaviour and learning difficulties (Woods, 1996).

Placental development
The development of the human placenta occurs in three waves–the first at ∼9—12 weeks, the second at ∼16 weeks and the third at ∼32 week gestation (Castellucci et al., 1990). It is, however, the first wave of trophoblastic invasion of the spiral arteries that appears to be the major physiological transition in placental development (Burton et al., 1999). Levels of FAAH in both the human placenta and the maternal circulation increase towards the end of the first trimester of pregnancy, before declining by the early second trimester (Maccarrone et al., 2000b; Helliwell et al., 2004). Furthermore, high levels of FAAH have been observed in the villous cytotrophoblast. Its expression in the syncytiotrophoblast suggest that FAAH in these cells help prevent the transfer of AEA from maternal blood (Helliwell et al., 2004), which, to some degree, has been suggested as a protective mechanism for the developing fetus (Helliwell et al., 2004). Previous studies have suggested that circulating FAAH and AEA levels may be critical to the outcome of early pregnancy (Maccarrone et al., 2000d, 2002b). It has been shown that decreased expression and activity of FAAH in peripheral blood lymphocytes is an early marker of early spontaneous abortion (Maccarrone et al., 2000d, 2002b). The importance of FAAH in early placental development has been supported by a recent study that demonstrated the expression of FAAH and CB2 (rather than CB1) receptors in the human first trimester placenta. This study also provides evidence that the endocannabinoid regulation within placental tissue is independent of the maternal immune system (Helliwell et al., 2004).

The mechanism whereby the main psychoactive ingredient of marijuana, Δ9-THC, causes FGR has been postulated to occur through four mechanisms. First, Δ9-THC crosses the placenta to a greater extent during the early proliferative growth phase compared with the hypertrophic growth phase of late pregnancy to affect the fetus (Vardaris et al., 1976). Secondly, the extended half-life of Δ9-THC in the maternal circulation results in prolonged fetal exposure (Hutchings et al., 1989). Thirdly, the indirect exposure of the fetus and the trophoblast to increased levels of carbon monoxide from smoking marijuana is directly toxic (Clapp et al., 1987), and fourthly, marijuana has a tendency to increase maternal heart rate and blood pressure (Sidney, 2002), inducing uterine vasoconstriction and thereby reducing feto-placental perfusion (Zuckerman et al., 1989). Recent evidence from our group demonstrating that Δ9-THC acts directly on the human cytotrophoblast cell to inhibit cell growth and the transcription of genes involved in growth and apoptosis (Khare et al., 2006), via the relatively unstudied CB2 receptor (Taylor et al., 2007) suggest that endocannabinoids may similarly inhibit trophoblast growth and development.

Vascular effects: potential mechanism for reproductive failure

One possible factor, which has been shown to have direct effects on the vasculature and could be a contributing factor to the disruption of a low pressure, high volume flow placental system so necessary in normal pregnancy, is AEA. This speculation is based on ex vivo studies showing that AEA produces complex and variable vascular effects (Plane et al., 1997; Randall and Kendall, 1998), depending on the species and vascular bed studied (Pratt et al., 1998; Chaytor et al., 1999), with both endothelium-dependent and endothelium-independent vasorelaxation that may involve several different mechanisms (Fig. 4) (Randall et al., 1996; Chaytor et al., 1999; Wagner et al., 1999; Zygmunt et al., 1999).

AEA has been shown to have pleiotropic effects on the cardiovascular system in vivo. In anaesthetized rats, AEA causes a brief pressor and more prolonged depressor effect (Varga et al., 1995). In humans, both marijuana smoking and the intravenous administration of Δ9-THC have resulted in peripheral vasodilation and tachycardia (Dewey, 1986). These effects manifest themselves as a fall in peripheral resistance and hence blood pressure. The control of resistance within the vasculature is important because it ultimately affects blood flow and tissue/organ perfusion, and this has major implications for normal placental development and trophoblast survival (Battaglia and Regnault, 2001; Haggarty, 2002; Illsley, 2002).

Endothelium-dependent responses
Several groups have reported a vasodilatory effect of AEA through the release of various endothelium-derived releasing factors (Randall and Kendall, 1998; Harris et al., 2002). Using the rat model, Deutsch et al. (1992) showed that exposing cultured renal endothelial cells to AEA stimulated the release of nitric oxide, while the presence of the nitric oxide synthase inhibitor, L-nitroarginine methyl ester (L-NAME), abolished this response.

Another endothelium-dependent vasodilator, the putative endothelium-derived hyperpolarizing factor (EDHF; Fig. 4), has been suggested as one of the mechanisms of AEA-induced vasorelaxation (Randall and Kendall, 1998; Harris et al., 1999; Jarai et al., 1999), whereas others have demonstrated that AEA induces membrane hyperpolarization via another unknown intermediate (Zygmunt et al., 1997). It has also been suggested that in rabbit mesenteric arteries, the endothelial component of AEA-induced vascular relaxation is secondary to gap junction communication and not to a specific vasoactive factor (Chaytor et al., 1999). Therefore, the role of EDHF in AEA-induced relaxation may vary according to the tissue species and vascular bed studied.

Endothelial (HUVEC) cells also express the CB1 receptor and the presence of these receptors are postulated to confer an anti-apoptotic effect via an AEA-dependent mechanism (Maccarrone et al., 2000c; Yamaji et al., 2003). These receptors have been localized by immunohistochemistry in the endothelium and trophoblast cells of human term placenta and gestational membranes (Park et al., 2003), which these authors suggest could be involved in the development of FGR, whereby the placental expression of CB1 receptors may be modified, leading to varying degrees of apoptosis and FGR. Importantly, the presence of CB1 receptors in the placenta and fetal membranes at term may be related to parturition, although these data require further proof.

Endothelium-independent responses
AEA has also been shown to exert some of its effects directly on vascular smooth muscle via the CB1 receptor, independent of the endothelium (Gebremedhin et al., 1999). The contribution of direct effects on smooth muscle to the vasodilatory effects of cannabinoids varies among vascular beds, as does the mechanism of action (Hillard, 2000). For example, AEA has been shown to inhibit the opening of L-type calcium channels (Gebremedhin et al., 1999), thereby modulating calcium influx (Ho and Hiley 2003a), release (Fimiani et al., 1999) and sensitivity. Other studies have demonstrated AEA-induced inhibition of smooth muscle intracellular calcium stores (Zygmunt et al., 1997), as well as modulating calcium entry through voltage-gated calcium channels (Gebremedhin et al., 1999; Ho and Hiley, 2003b).

Other endothelium-independent mechanisms responsible for the vasodilatory effects of AEA include the release of calcitonin gene-related peptide (CGRP), a vasodilator released from perivascular sensory nerves, and activation of CGRP receptors on vascular smooth muscle by interacting with VR1 receptors on rat and guinea pig perivascular sensory nerve endings (Zygmunt et al., 1999; Mukhopadhyay et al., 2002). Vasorelaxation induced by CGRP has been proposed to be mediated by two pharmacological actions, one being the stimulation of smooth muscle cell adenylate cyclase with the subsequent accumulation of intracellular cyclic adenosine monophosphate (cAMP) (Edwards et al., 1991), and the other, the activation of potassium channels within the vascular smooth muscle (Brayden et al., 1991), which is known to cause smooth muscle cell relaxation and inhibition of agonist-induced contractions (Wray et al., 2003).

It is possible, therefore, that arterial and venule endothelial cells which may have different receptors, different signalling pathways and different phenotypes will exhibit differential responses to AEA.

Hormonal regulation of AEA-induced vasorelaxation
At physiological concentrations, estrogen stimulates the release, rather than the uptake of AEA from endothelial cells, leading to a rapid elevation of intracellular calcium and nitric oxide release (Maccarrone et al., 2000a). In addition, 17 β-E2 has also been shown to increase vascular sensitivity to CGRP (Gangula et al., 1999), suggesting that estrogen modulates the vascular effects of AEA. However, other studies have found no sex-linked difference in the vasorelaxant effects of AEA (McCulloch and Randall 1998). There appears to be little else published on the effect of gonadal or adrenal steroids on vascular smooth muscle cell and cannabinoid interactions.


The detrimental effects of AEA on the embryo, the oviduct and the blastocyst are prevented by the presence of an efficient system for its removal which is under hormonal control, showing an interplay between endocannabinoids and gonadal steroid hormones in regulating fertility in mammals (Maccarrone et al., 2000b).

The expression of CB1 receptors in the pre-implantation embryo, oviduct, uterus and sperm and the synthesis of AEA in the pregnant uterus and sperm suggest that cannabinoids may be involved in the regulation of all aspects of reproduction including fertilization, pre-implantation embryo development, implantation and placental development. These data have led to the hypothesis that a major imbalance in FAAH or AEA levels at any stage of reproduction may lead to reproductive failure. Indeed, as the placenta develops at the implantation site, the amount of AEA required must be tightly regulated or implantation does not occur. Later imbalances in AEA or FAAH levels lead to spontaneous abortion, and thus it is possible that when the disruptions are not severe enough to result in spontaneous abortion, trophoblast development and interaction with endothelial cells becomes unstable, ultimately impairing placental function. Even transient impairment could be sufficient to result in placental dysfunction. All these data implicate cannabinoid signalling in normal vascular or trophoblast responses that lead to normal pregnancy, and any minor disruption may result in complications of pregnancy related to placental dysfunction such as FGR and pre-eclampsia to FGR, with any major disruption leading to spontaneous abortion. Table 1 provides some supporting evidence for the effects of cannabinoids on reproductive outcomes in the human and some areas where they are speculated to have an action. The table also points to areas of future research, where the effects of these compounds are currently unknown and where further work is required in order to validate these hypotheses in the human.

Source, Graphs and Figures: The role of the endocannabinoid system in gametogenesis, implantation and early pregnancy